Minimum qualifications:

• Bachelor’s degree or equivalent practical experience.
• 5 years of experience with software development in Python.
• 3 years of experience testing, maintaining, or launching software products, and 1 year of experience with software design and architecture.
• Experience with compiler optimization, code generation, and runtime systems for popular accelerators, including GPU or TPU.
• Experience with Machine Learning, Machine Learning Optimization, Performance Optimization, and Large Language Models.

Preferred qualifications:

• Master's degree or PhD in Computer Science or related technical field.
• Experience with TPU architectures, memory hierarchies, performance bottlenecks, and Accelerated Linear Algebra.
• Experience tailoring algorithms and ML models to exploit TPU architecture strengths and minimize weaknesses.
• Experience with various stack layers: compilers (OpenXLA, MLIR), serving libraries/frameworks (vLLM, sglang) and ML frameworks (JAX, PyTorch).
• Debugging experience to improve performance of single-mode or multi-mode (distributed) systems.

About the job

Members of the team work on research and develop audio representations that best captures both the semantic and acoustic information within the audio signal to enable Gemini models to understand audio inputs and also generates natural audio sounds. This builds the foundation for audio-to-audio dialog models, text-to-speech, speech-to-speech translation, duplex audio dialogs.

Artificial intelligence will be one of humanity’s most transformative inventions. At Google DeepMind, we are a pioneering AI lab with exceptional interdisciplinary teams focused on advancing AI development to solve complex global challenges and accelerate high-quality product innovation for billions of users. We use our technologies for widespread public benefit and scientific discovery, ensuring safety and ethics are always our highest priority.

Responsibilities

• Design, build, and maintain training infrastructure to support Gemini Audio encoder and pretraining models, focusing on Accelerated Linear Algebra (XLA) optimization for training on Tensor Processing Units (TPUs).
• Implement tools to analyze and track audio model training efficiency and health.
• Contribute and work closely with Gemini to build and maintain audio related components.
  
• Collaborate with research teams across Gemini Audio to improve the model training and inference efficiency.